TSMC is ramping up capacity at its 2nm-class process node as orders from Apple and Nvidia increase, according to a TipRanks report.

The world's largest contract chipmaker has seen demand for its N2 technology strengthen, with Apple and Nvidia emerging as the leading customers driving the capacity expansion. Apple has historically been first in line for each of TSMC's leading-edge nodes, and the report indicates that pattern continues with the transition to 2nm.

Nvidia's inclusion among the early 2nm customers marks a shift in TSMC's customer mix at the leading edge. The AI chip designer has become one of TSMC's largest revenue contributors over the past two years as data center demand for its GPUs surged. Its commitment to N2 deepens that relationship.

TSMC began high-volume production on N2 in late 2025 at its fab in Hsinchu, Taiwan. The node uses gate-all-around nanosheet transistors, replacing the finFET architecture TSMC has employed since 16nm. The company has said N2 will deliver meaningful gains in performance and power efficiency over the preceding N3 and N3E generations.

Ramping a new node is capital-intensive, and TSMC has been investing accordingly. The company's annual capital expenditure has run in the range of roughly $30 billion in recent years, with a substantial share directed at leading-edge capacity in Taiwan and at its new overseas sites in Arizona and Japan.

The order momentum from Apple and Nvidia gives TSMC visibility into N2 utilization as it climbs the yield curve in the early quarters of production. Strong anchor customers typically help absorb initial wafer output while yields mature, smoothing the economics of the ramp.

For Apple, 2nm chips are expected to power future iPhone and Mac silicon, continuing its strategy of pairing custom A-series and M-series processors with TSMC's most advanced nodes. For Nvidia, the node offers a path to denser and more power-efficient AI accelerators as competition in the AI chip market intensifies.

TSMC's rivals are moving on similar timelines. Samsung Foundry is developing its own 2nm-class technology, and Intel is advancing its 18A node, but TSMC retains the largest share of the foundry market and the deepest roster of leading-edge customers.

The report does not specify exact wafer allocation figures or revenue projections tied to the Apple and Nvidia orders, and TSMC has not publicly broken out N2 capacity by customer. Analysts broadly expect leading-edge demand tied to AI hardware to keep TSMC's advanced nodes sold out through the coming cycles.

If Apple and Nvidia order flow holds at the pace described, TSMC's N2 capacity should tighten quickly, reinforcing both pricing power at the leading edge and the company's advantage over rival foundries through 2026.